Blasting 1 Million Tons, 205 Meters From a Town

Dr Thierry Bernard, DNA-Blast &

Guy Gagnon, OSISKO Mining Corporation

Abstract

Blasting at very close proximity to urban areas is typically a situation where both local communities and

mining stakeholders get nervous. The context becomes logically even more tense when the blasting

operations have to be performed above an old underground mine. When mitigating subsidence risks

needs the largest possible shots to be fired, the nuisance control (flyrock, vibration, and noise) required

by local and national regulations, contrarily urge for being more conservative.

OSISKO Mining Corporation is a Québec-based mining company whose main property, 100% owned, is

located in the heart of the prolific Abitibi Gold Belt in Quebec, immediately south of the town of

Malartic, approximately 20 kilometers west of the town of Val d'Or. As of January 1, 2013, the open pit

Proven & Probable gold reserves stand at 10.1 million ounces. Since the beginning of operations in

2011, the total production amounts to 588,615 ounces of gold. The property includes the former

Canadian Malartic underground mine, which produced more than 1 million ounces of gold from 1935 to

1965.

This papers presents OSISKO’s experience in simulating, designing and shooting large to very large-

scale blasts that conciliates the company’s commitment to sustainable development, the strict

observance of local and national legislation, and the prevention of ground movement that could affect

the achievement of the operational objectives.

Particular emphasis will be on integrating innovative blasting digital simulation techniques into the daily

operational process, offering OSISKO the opportunity of shooting a 1 million ton successful mega-blast,

205 meters from the closest residential area.

Location and Geology of the Malartic Site The Canadian Malartic gold-bearing deposit is located in Québec, in Canada, just to the south of the

town of Malartic, approximately 20 km to the west of the town of Val-d’Or. Its coordinates are 48º 7’

45” latitude N and 78º 7’ 00’’ longitude W. The deposit overlaps the southern boundary of the eastern

part of the Abitibi gold belt that contains many large gold-bearing deposits. The exploration phase began

in 2005. The mine was developed in just 6 years. The first gold ingot was cast in April 2011, and

commercial production began in May 2011.

The deposit is an archean porphyritic gold-bearing system, consisting of a widespread disseminated ore

halo of gold and pyrites, within a dioritic porphyry and altered meta-sediments. Drilling and data

collection defined the boundaries of a gold-bearing system measuring 1900 meters x 350 meters, with a

real thickness varying from between 40 to 270 meters, to a vertical depth, from the surface, of 320

meters. The deposit is composed of 10.1 million ounces of proven and probable reserves (310.6 Mt @

1.01 g/t Au, according to rule 43-101), 11.70 M oz of overall measured and indicated resources (347.3

Mt @ 1.05 g/t Au), and 1.20 M oz of supposed overall resources (49.6 Mt @ 0.75 g/t Au).

The feasibility study of November 2008, as well as the following updates of reserves in February 2010,

March 2011 and February 2013 revealed a potential open-pit production of between 500 000 and

600 000 ounces of gold per year, for a 16-year operation, with a daily processing capacity of 55 000

tons. Ore processing is via a traditional cyanuration process, then the precipitate is extracted by

absorption on activated carbon (CIP or "carbon-in-pulp" process). The Canadian Malartic mine required

investment in excess of $ 1 billion.

Specific Site Constraints

The property includes the old Canadian Malartic underground mine that produced over one million

ounces of gold from 1935 to 1965, from ore whose content varied from 3 to 6 grams per ton Au, and

over 5 million ounces, if we count from 1935 to 1983. The area worked is partly situated above the old

galleries of the underground mine, thus creating significant subsidence risks. The presence of these

galleries has a direct impact on the operating methods and safety measures implemented to protect

persons and property.

The mine (figure 1) reaches as far as the town, half of which has been relocated to the north (figure 2), at

a cost of $ 160 million after the exploration phase. Therefore, this is a mine "in a town", with all the

associated constraints: flyrock, vibration, noise, dust and blasting fumes.

Figure 1: view of the mine close to Malartic town

Figure 2: OSISKO’s resettlement program for part of the town of Malartic (2008-2009). The

relocated southern district, partly closed due to subsidence risk, is outlined on the left-hand side,

while the new district appears on the right.

Several measures were taken to reduce the impact of mining on the town of Malartic The drilling

equipment was soundproofed, matting was added to the truck skips, and the position of the plant,

crushers and conveyors was optimized to decrease the noise impact of the project.

A huge protective so-called "green wall" (figure 3 and 4) separates the pit from the town. A few meters

behind the wall there is a park with a children's playground, shops and housing.

Figure 3: Location of the "Green wall" and of the seismic monitoring points

Figure 4: Detail view of the "Green Wall"

Site Operating Principles Traditional open-pit mining methods were chosen for the Canadian Malartic deposit because of its low

content and surface proximity.

Due to the large quantity of ore with a low content (estimated waste/ore ratio of 2.3:1), a high

production rate for the plant of 55 000 t/day, was defined as being profitable from an economic

standpoint, and would enable the value of the deposit to be maximized.

The result is a very conventional operating process, with stripping, then blasting phases in terraces

dropping to the pit. The waste and ore are loaded by three 28 m3 electric hydraulic excavators (O&K

RH340-B), three wheeled loading shovel (two LeTourneau L-1850 and one CAT 994). Transport is via

CAT 793F - 227 ton - rigid trucks.

General Blasting Constraints Blasts must be in compliance with the operating constraints (size distribution) but above all with the

environmental constraints due the proximity of the town.

No flyrock beyond the confines of the pit

Vibration (PPV <12.7 mm/s)

Overpressure (<128 dB)

Blasting fumes (none over the town)

No blasting is carried out when the wind is from the south in order to protect the town from blasting

fumes.

Blasts are scheduled at the least sensitive times of the day, and there is constant communication and

information with the authorities and the town's inhabitants.

Seven seismographs have been installed in the town of Malartic to measure the overpressure and

vibration levels of each blast (figure 3, here above).

The authorization of "special blasts", due to their being close to the town, their duration and/or size, are

managed individually, directly coordinated with the Ministry of the Environment of Québec and the

town of Malartic, and the monitoring committee including the residents. A decree of the Ministry of the

Environment enabled the OSISKO mining company to carry out blasts of up to 15 seconds. This

restriction was exceptionally extended to 37 seconds to satisfy the constraints for "special blasts"

scheduled for October 2012, also referred to as "mega-blasts" due to their unusual size. For blasts such

as these, in addition to the extensive use of blasting mats and sand, evacuation areas from 100 to 600

meters were defined. The authorizations had to be obtained individually with full and accurate

documentation. Recourse to digital simulation was proved to be decisive in obtaining the authorizations,

particularly thanks to the credibility that their scientific features offer.

General Blasting Design At the beginning of the project, a traditional design was used. A campaign of test blasts was undertaken

with small charges in order to define the seismic behavioral laws (K and alpha), and for the zoning of

the site, based on the maximum charges per delay to adhere to the seismic levels.

For each area, a diameter was also calculated, that was compatible with the charge per delay, taking into

account the flyrock risks. After an analysis, a diameter of 225mm was retained for the areas farthest

from the town, whilst a diameter of 89 mm was preferred in the areas closest to the residential areas.

The number of charges recommended per hole after the digital simulations varied from several, close to

the town, to one, further afield. The use of electronic detonators was generalized at the site to guarantee

compliance with the charge per delay, due to the size of the blasts (several hundred holes). The height of

the terraces was on average 10 m.

Specific Blasting Design In areas close to the town, operations were above the old underground mine. A network of galleries and

cavities transformed the solid rock into a "Swiss cheese" whose stability was difficult to assess,

particularly when very heavy vehicles drove constantly overhead.

For safety reasons, the mine therefore decided to mass blast these areas, in order to limit the risks that

would appear when making smaller blasts (decompression of land, uncontrolled subsidence when

digging or drilling for instance). This technique is similar to block caving. "It is a special works site. We

are above three open works sites that are interconnected. For the safety of our employees, we have to

blast in one go, to avoid subsidence", said Hélène Thibault, OSISKO's Communications Director. The

company's modeling led people to believe that, during the blast, the rock would fill the voids of the sites

below and stabilize the whole area.

Although this option was reassuring for the stability of the ground, it was worrying for the neighboring

population, for whom a large-scale blast (e.g. 1 million tons), represented a danger both for people and

housing. Therefore, in this configuration, blasting required a specific design and measures in order to

guarantee suitable and safe results, especially as far flyrocks, fumes and .

Specific Design for Flyrock Substantial final stemming (from 3.6 at 5 m with a diameter of 140 mm) was used to contain vertical

flyrock. In compliance with the site's practices, particularly in areas closest to the town, the blasts were

covered with a layer of inert material (like sand) or matting to limit and contain any stemming ejection

or cratering.

Specific Design for Fumes Blasting fumes (orange fumes) that can sometimes be perceived at a blast are the result of poor

detonation of the explosive. This can have several origins; but in the present configuration, it can be

specifically due to the effect of time on the explosive. Due to the size of the blasts, loading takes several

weeks. Water can alter the explosive or cause some holes to collapse, leading to dilution via the loss of

the confinement. The latter can occur during the dynamic phase of the blast via decompression in the

former galleries or chambers.

The measures taken to limit the exposure to these harmful blasting fumes consist of only initiating blasts

when the prevailing winds are favorable (winds from the north), blowing any potential fumes in the

opposite direction to the inhabited areas. These measures are not specific to "mega-blasts", but their

criticality increases with the size of the blast and its location above the underground mine.

Specific Design for Vibrations The first measure, described above, is to limit the charge per delay, according to the different areas of

the pit. The method used is a traditional method based on the calculation of an attenuation law (K and

alpha parameters). This method leads to using decked charges closest to the town.

The second measure consists of using the "Signature Hole" technique which allows us to oppose the

phases of the seismic waves from each of the charges. This technique is based on blasting single holes

(one single charge) for which the vibrations are measured at the critical points (housing), and the use of

electronic detonators enabling accurate firing at the desired time. Thus, optimized firing sequences are

calculated for each blast.

Characteristics of a Typical “Mega-Blast” In this context, large so-called "mega-blasts" are used. The example below describes the typical

characteristics of one of these mega-blasts.

Number of holes: 1025 (figure 5)

Number of charges: 1806 (figure 6)

Diameters: 115 and 140 mm

Charge per delay: 70 kg

Specific charge: 480 g/m3

Minimum Inter hole delay: 6 ms

Duration of the initiation sequence: 22.5 s (figure 7)

Average depth of the production holes: 10 m

Depth of the block caving holes: up to 50 m

Total quantity of explosive: 110 tons

Volume of the muck pile: 250 000 m3 (500 000 t)

Closest housing: 205 m

Figure 5: Two 3D views of the 1025 hole blast

Figure 6: Hole loading from single to multiple decks

Figure 7: Final sequence for Mega-Blast #2

It is easy to understand that this type of blast could alarm the neighboring population. During one of

these operations, on 27 October 2012, the Citoyens du Quartier-Sud de Malartic residential group said

they would keep a close eye on this "special blast," as it had been defined by the company. Their

spokesperson, Carl-Hugues Leblanc was worried about the proximity of the inhabited area. "Some

dynamiting will be only 70 feet from the residence of one of our members. So we are very worried," he

told a reporter sent to the site by Radio Canada.

Indeed, the OSISKO mining company was preparing to blow up 940 000 tons of waste and ore. Months

of preparation were necessary. The importance of the communication and information plan managed by

OSISKO was in the case in point of primary importance.

A safety zone was set up and thirty or so company’s employees were positioned to prevent people

accessing dangerous areas, should any flyrock be cast. The residents closest to the blasting area were

evacuated as necessary as a precautionary measure. As early as mid-September, the company sent out a

letter to each resident in the town informing them that larger, and above all, longer blasts than those

usually made would take place "shortly."

Blast mats were positioned and a thick layer of sand was spread over the surface bench. The sequence

initially scheduled to last 37 seconds was divided into two separate operations in the end: one lasting

22 seconds, shortly followed by another of 15 seconds. The usual length of a blast at Malartic was four

to six seconds.

The Dispensatory Nature of "Mega-Blasts" The Minister of the Environment of Québec, Daniel Breton, had given his consent for a 37-second blast

subject to conditions; in particular requiring the setting up of a safety zone to protect the site's

employees, but also the formal request of an independent study, with an expert's report to be submitted

to the ministry.

Thus, the authorization was granted via a dispensation from the Ministry of the Environment of Quebec

to whom OSISKO had to provide the following:-

A report on the simulation of flyrock for the blasts

A plan of the safety zone implemented

A plan of the safety measures implemented to protect persons and property

It must be stressed that this operation was in the media limelight, in particular due to the unusual

immediate proximity of housing to this mega-blast, and the political context surrounding the

dispensatory procedure.

How to Guarantee the Results of Such a Blast An explanation of all the technical measures (state of the art) taken by a team of professionals to carry

out a blast such as this correctly, was not sufficient to convince people of their being well founded. The

size of the blast (in terms of the rock blasted, the quantity of explosive), and the proximity are

frightening!

"We informed the residents to avoid them being surprised, but we carried out tests and simulations in

order to comply with the vibration and overpressure figures," Hélène Thibault asserted a few days before

the blast. The simulations were at the core of the arrangements, and enabled us to obtain the necessary

administrative authorizations, whilst providing the different stakeholders with an additional degree of

confidence.

In order to provide the necessary guarantees, digital simulation was used: a technique that consists of

virtually reproducing the effects of a blast to make sure that the result will be conform to the objectives,

be they for risks and nuisances (flyrock, vibration, noise) or the size distribution of the blast.

A Model Based on Physical Equations To simulate is to forecast. In this respect, two types of model are in direct opposition. Models based on

statistics allow inputting different configurations and results, and the model can interpolate for other

configurations. This type of model is limited in that it is only capable of simulating within the scope of

the reference cases.

Now in our "mega-blast" situation, we are confronted with new relatively unique situations (the first

blast takes place without there being a comparable operation to serve as a reference). Therefore, we have

to turn to the second type of models: models based on physical principles. The advantage is that they

only require one reference case to be operational (calibration). The universality of the physical principles

allows all types of configurations to be calculated, with an excellent degree of reliability, even if they are

far-removed from the reference case. This is the case of the DNA-Blast model and its I-Blast software.

Therefore, this is the model (figure 8) that was chosen and used to simulate the effects of these mega-

blasts.

Figure 8: Vibration simulations on a 3D view of the blast: seismic waves in red

Signature Holes and Digital Simulation of Vibration The Signature Hole principle was used, as previously mentioned. In this configuration, the use of

physical principles for digital simulation is very relevant. The model must be capable of simulating the

seismic signal, its maximum level of vibration and its frequency content, for more than 1800 spaced out

charges. The presence of decked charges also requires the simulation technology to take into account the

3 dimensions in its representation, and more importantly in its calculations.

With 1806 charges and 1025 holes, the hypothesis that the signal of a signature hole would be

representative of all the holes in the volley can appear to be risky. This is why six signature holes (figure

9) were made, separated by 1 second and located as close as possible to the area to be mined. Six of the

single holes gave a workable waveform.

Figure 9: Location of the 6 signature holes (red squares), at left to the Mega-Blast

The results showed different amplitudes (figure 10) and wave shapes, but a relatively similar frequency

content (figure 12). Eight simulations corresponding to the eight signature holes were carried out, which

enabled the confinement differences to be taken into account and provided an estimated range of the

vibration levels.

Figure 10: Seismic traces of the 6 signature holes

An attenuation law (figure 11) was performed on the signature hole PPVs recorded at various distances

(figure 12) in order to feed the simulation model. Results were very consistent, building confidence in

the simulation results.

Figure 11 : Attenuation law based on signature holes

Figure 12 : Typical frequency content of a Signature Hole

For reasonably sized blasts, the firing sequence was optimized using the signature hole method (figure

13).

Figure 13 : Optimum delay as a result of processing 6 SH for one location

However, in the case of a mega-blast, the sequence was constrained, firstly by the configuration of the

blast itself, and the limit of the programmable delay range of electronic detonators.

In the case in point, electronic detonators enabled us to have 22 000 ms delay available. Consequently

the mega blast was split into two mega-blasts respectively of 1025 and 790 holes. For mega blast #1, as

an example, 1025 holes represent 1806 charges: this provided a maximum of 12 ms between charges.

The firing order of the charges was defined by the practitioner (expert) according to his knowledge of

the underlying ground. The first simulation (figure 14) was then carried out with a 8 ms inter charge

delay and showed a particle velocity (PPV=19.5 mm/s) that was not compatible with the vibratory limit

of 12.7 mm/s.

Figure 14 : Vibration simulation with 8 ms Inter Charge delay

A virtual trial and error process was then applied thanks to the digital simulation software in order to

discover, via the iterations, the deviation between charges that would be compatible both with the

structural limit of the blasting accessories, and that would satisfy the limit in terms of PPV. For final

design, delays between charges of 10 and 12 ms were used (figure 15). The simulated PPV was 5.9

mm/s.

Figure 15 : Vibration simulation with 10-12 ms Inter Charge delay

Flyrock Simulation A flyrock simulation was also undertaken for each of the holes with a free face, in order to predict the

maximum horizontal projection distance. A simulation of the other blast holes was carried out to

analyses the location and magnitude of any vertical flyrock, particularly caused by stemming. Each hole

was accurately recorded (final stemming, depth of cover (figure 16), so the simulation (figure 17)

underlined the boundary conditions that were then corrected on site.

Figure 16 : stemming depth reporting for each hole

Figure 17 : Fly rock simulation depending burden

Presentation of the Simulation Results In the case in point, communication and information were crucial, as we saw. The presentation of the

results in the form of graphs showing the ranges of incertitude enabled a better account to be taken of the

simulation results to implement the relevant recommendations and safety measures.

Both from the standpoint of administrative authorizations as well as the debate with the political world,

or residents groups, the clear and exhaustive presentation of credible simulations, based on physical

principles enabled a certain degree of serenity to be maintained and helped with a favorable outcome for

this operation.

The Mega-Blast Design and Implementation The implementation of the Mega-Blast ran through particularly challenging constraints, starting from

ground instability due to the already mentioned old galleries. Drilling and loading operations were

particularly affected as sown in figures 18 and 19.

Figure 18 : Drilling safely over an instable area

Figure 19 : Loading hole safely over instable area

The high number of charges created a dense network on the surface (Figure 20), requiring extensive

craning to allow the planned sand-covering stage without damaging the detonators cables (Figure 21).

Figure 20 : EDD wiring network

Figure 21: Covering the blast area with sand (craning to avoid damaging the detonator cables)

Similarly, the areas situated at the closest distance between the mega-blast and the residential area, were

covered with a very large number of blast mats, as shown in figure (Figure 22).

Figure 22 : Mat covering on closest holes to the town

Finally, a specific evacuation area was defined and implemented (Figure 23)

Figure 23 : Evacuation area

The Mega-Blast: the Results The seismic results were conforming to the forecasts of 4.4 mm/s for a simulated range of between 3 and

5.5 mm/s (figure 24).

Figure 24 : PPV simulated versus actual

The frequency content and the seismic trace were also in line with the forecasts (figure 25)

Figure 25 : Comparing predicted (blue) and measured (green) waveform

All the face flyrock was very close to forecasts as well as the ejection of the stemming. Only one area

seriously decompressed into a cavity and created a vertical ejection (figure 26) causing a few fragments

to fall within the confines of the safety zone. Today, this type of unforeseen circumstance is difficult to

take into account and represents the limit of the simulation exercise.

Figure 26 : view from the town of the exceptional blast carried out on October 27 at Malartic mine

Photo: Daniel Rompré / Ville de Malartic

Conclusion Mega-blasting on the site at Malartic, very close to the town, was possible, adhering to the various

constraints be they, seismic, flyrock, or blasting fumes, thanks to the perfect control of the blasting

parameters, the use of modern technology such as electronic detonators and digital simulation.

Digital simulation based on physical principles offers:

The possibility of finding the best set of parameters thanks to virtual trial and error

An additional degree of confidence, in the choice of the blasting parameters when the blast is very

complicated or exceptional.

However, although unforeseen circumstances, mainly geological, are substantially reduced, we must

remain humble and take all the necessary precautions and protection with regard to persons and

property, because unforeseen circumstances can never be reduced to zero.